Abstract
We introduce an iterative method to search for time-optimal Hamiltonians that drive a quantum system between two arbitrary, and in general mixed, quantum states. The method is based on the idea of progressively improving the efficiency of an initial, randomly chosen Hamiltonian, by reducing its components that do not actively contribute to driving the system. We show that our method converges rapidly even for large-dimensional systems and that its solutions saturate any attainable bound for the minimal time of evolution. We provide a rigorous geometric interpretation of the iterative method by exploiting an isomorphism between geometric phases acquired by the system along a path and the Hamiltonian that generates it. Our method is directly applicable as a powerful tool for state preparation and gate design problems.